Esecurityplanet
OpenClaw Vulnerability Allows AI Log Poisoning via Unsanitized Headers
First seen 18 Feb 2026, 18:23 UTC
•
•29.2
Export
Article Content
Browse articles
OpenClaw versions prior to 2026.2.13 have a vulnerability that logs unsanitized WebSocket headers. This flaw creates a risk of AI log poisoning, potentially affecting systems that utilize these versions. Users are advised to update to the latest version to mitigate this risk.
Analyzing cluster data...
Referenced clusters:
Something went wrong. Please try again.
Timeline
2026-02-17
Reddit post about log poisoning in OpenClaw
2026-02-18
Esecurityplanet article published detailing the vulnerability
More articles in this cluster
Continue Reading
Iranian APT Groups Target Israeli Organizations with Modular C2 Frameworks
Critical Vulnerability in Cline Kanban Exposes AI Coding Agents to Hijacking
FBI Warns of Kali365 Phishing Kit Targeting Microsoft 365 Users
Mirax Android RAT Transforms Infected Devices into Proxy Nodes
ZeroTokens Phishing Platform Enables Real-Time Attacks on Financial Institutions
Surge in Android Banking Trojans: Four Campaigns Identified